|
Soniya Kharra
Keywords:
Opinion Mining, Machine Learning, Healthcare Feedback Analysis, Sentiment Analysis, Patient Satisfaction, TF-IDF, Support Vector Machine, Text Classification.
Abstract:
The growing use of digital healthcare platforms has increased the availability of patient feedback, creating opportunities to assess satisfaction and identify areas for service improvement. This study presents a machine learning-based opinion mining framework for classifying healthcare feedback into positive, negative, and neutral sentiments. The methodology involves text preprocessing, including data cleaning, tokenization, stop-word handling, and text normalization, followed by feature extraction using Term Frequency–Inverse Document Frequency (TF-IDF). Four supervised machine learning algorithms—Logistic Regression, Linear Support Vector Machine (SVM), Multinomial Naïve Bayes, and Random Forest—are comparatively evaluated using accuracy, precision, recall, macro F1-score, and confusion matrix analysis. The tentative results indicate that Linear SVM achieves the highest performance, with 91.6% accuracy and a 91.0% macro F1-score, compared with the other evaluated classifiers. The framework also considers feedback themes such as staff behaviour, waiting time, cleanliness, communication, and facilities to identify potential areas of patient satisfaction and dissatisfaction. The proposed approach demonstrates the potential of TF-IDF-based machine learning techniques to support systematic healthcare feedback analysis and data-driven service improvement. However, the reported performance metrics and thematic findings are illustrative and must be validated using actual experimental data before being reported as final research findings.
|
|
International Journal of Recent Research and Review
ISSN: 2277-8322
Vol. XIX, Issue 3
September 2026
|
PDF View
PUBLISHED
September 2026
ISSUE
Vol. XIX, Issue 3
SECTION
Articles
|
